C++ version of openWakeWord.
- Download a release of the onnxruntime and extract to
lib/<arch>where<arch>isuname -m. - Run
make
After building, run:
arecord -r 16000 -c 1 -f S16_LE -t raw - | \
build/openwakeword --model models/alexa_v0.1.onnxYou can add multiple --model <path> arguments. See --help for more options.